OpenAI is extending access to its DALL-E 3 image generation model to free-tier users. Previously available only to ChatGPT Plus subscribers, this model now allows free users to generate up to two images per day. The announcement was made via a post on social media platform X, where OpenAI highlighted that users can create images based on their prompts through ChatGPT.
DALL-E 3, launched in September of the previous year, enables users to generate detailed images from textual descriptions, thanks to its advanced conversational interface. This feature simplifies the process of creating precise image prompts.
In addition to generating images, OpenAI has introduced an editing feature for DALL-E 3. This tool allows users to select and modify specific areas of generated images by providing text-based descriptions of desired changes.
In other news, OpenAI has released the “System Card” for its latest multimodal AI model, GPT-4o. This document outlines the safety measures and risk assessments associated with the model. It reports that GPT-4o has been rated as “Medium Risk” overall, with evaluations in categories such as cybersecurity, biological threats, model autonomy, and persuasion. Although the model was rated low risk in most categories, concerns were noted about its potential for generating persuasive content.
